Key words: Alcohol dependants; Raven's Coloured Matrices; wrong answers; analysis; identity by contiguity; opposition; confabulation; neuronal mechanisms.

This analysis in demented (Pola, Cocchi and Zerbi, 1988), in college students (Cocchi, 1993) or in alcoholics (Cocchi, 1993, 1995; Cocchi and Meme' 1998) brought to light non random grouping of many wrong answers. According to these papers, the split of the prevalent mistakes shows the cognitive mechanisms in use to elicit these answers in a problem solving task.

Now I want to evaluate the wrong answers given by one group of alcoholics under detoxification who scored 20.

So because first paper (Cocchi 1993) did not collect data about alcoholics scoring 20, following a personal information then known as wrong. The score range of subjects with diagnosis of dementia is 0-20 and not 0-19 as I first heard (Bertolani, De Renzi & Faglioni, 1993). These authors moreover classified people scoring 21-25 as dubiously demented and people scoring 26-36 as surely non demented.

Being the second paper about alcoholics scoring 21-25 (Cocchi 1995) and the third one about those scoring 26-35 (Cocchi and Meme', 1998), people scoring 20 did not get any analysis till now.

Each subject had her/his sex and age collected. Via DSM-IV (F10.24) everyone of

them met the diagnosis of alcoholic dependence. Each subject did RCM few days after admission, or when the symptoms of alcohol withdrawal cleared up.

RCM is a multiple choice test distributed into three sets of 12 tables (Series A, Ab

and B), each table with an incomplete matrix of relations.

To complete the pattern, a testee must select one of the six alternative answers set below the matrix, but five of them lead to a wrong result.

If we mark any possible answer with a number from one to six, from left to right, we have five figures to point the wrong answers up. The remaining one is the right answer, but its figure name varies in each table.

As for RCM scoring, a right answer makes one point, the maximum score being 36.

Only RCM charts with 20 points score fit this research.

We checked up the RCM charts where the psychologist noted the figure of the answer the subject chose.

We chose only 29 matrices out of 36 because the probands did not make any mistakes in the remaining ones.

From the series A we examined the matrices A5-A12; from the series Ab the matrices Ab3-Ab12, and matrices B2-B12 from the series B.

The wrong answers were counted out and grouped table by table using to the figure name of any wrong answer.

Then each table had all its wrong answers divided by five, being the ratio token in excess to the closer figure. In this way each table's group of wrong answers had its theoretical number of random wrong answers.

These ratio figure of random answers token away from every group of a table, every remaining group shows the total of surely non random wrong answers, with a significant minimum level of 3 wrong answers (p = 1/5 ^3 = .008).

Then, I divided these significant groups into four fields as it follows: Partial identity, identity by left side or up contiguity in the matrix, opposition of colour or shape, confabulation (when the answer has extraneous details).

Being 20 score the fugure that exactly interposes people scoring < 20 from those scoring 21-25 I should expect features common to both previous samples.

Results.

All the RCM charts of these 20 alcoholics (M = 14; F = 6; M/F = 233.33; average age = 46.35 +/- 10.43, range: 30-74) fit the above criterion.

No subject refused to answer to whatever table, so every RCM chart had the expected 36 answers.

Of 316 wrong answers, Table 1 shows their number for each RCM table divided into the figure name of chosen responses from matrix A5 to matrix B12.

RCM tables no. 1-4 of series A, no. 1-2 of series Ab and no. 1 of series B did not have any room in Table 1, because we found all their answers right.

Beside the column of each of the six figure names, there is a second column, (n)bis, which indicates the amounts of surely non random wrong answers, from a minimum group of three (3+).

Table 2 splits the wrong answers groupings according to suppesed mechanisms of neuronal network functioning like similarity, contiguity, and opposition. To them I added confabulation, because some wrong answers cannot otherwise be explained.

Tables 3-5 compare wrong answers groupings as found in present research with those reported in my previous ones.

Table 1: groups of wrong answers, matrix by matrix, and significant groups of non random wrong answers. The symbol [*] runs for the right answer.

Series A

Keys for p: + = .008; ! <.002; # <.0004; @ <.00007; & <.00002; $ <.0000006

As you can see, significang groupings of wrong answers are increasing from the series A to the series B.

Table 2: non random wrong answers divided into the four chosen categories.

-------------------------------------------------------------------------------------------------

RCM Tab.no./ identity contiguity opposition confabulation

answers no. answers answers answers answers

Totals 0 = 0.00% 74 = 83.15% 12 = 13.83% 3 = 3.37%

(*) contiguity by diagonal.

As we can see, all 89 non random wrong answers found room into these four fields.

Table 3: Comparison among significant groups of wrong answers, matrix by matrix,

as reported in 1993 (93), 1996 (96) and present paper (99) for series A.

As you can see, only the 7.4 grouping is the same in all three research, while 9.5; 8.6 and 10.6 were found only in <20 scoring

and 20 scoring alcoholics. I can suggest that this way of grouping belongs only to people scoring less than 21, and to surely

Table 4: Comparison among significant groups of wrong answers, matrix by matrix,

as reported in 1993 (93), 1996 (96) and present paper (99) for series Ab.

As you can see, only the 11.3 and 12.1 groupings are the same in all three research, while 8.3 and 10.6 were found only in 20 and 20-25 scoring alcoholics. I can suggest that this way of grouping belongs only to dubious demented people, according to Bertolani, De Renzi & Faglioni, 1993. On the other hand, groupings 6.2; 6.5; 7.2; 8.2, 9.3; 9.5; 10.2; 11.2; 12.4 found both to <20 and 21-25 scoring people probably lack the same groupings in the present sample because its small number of subjects.

Table 5 Comparison among significant groups of wrong answers, matrix by matrix, as reported in 1993 (93), 1996 (96) and present paper (99) for series B

As you can see, only the 3.2; 5.5; 6.2; 7.2; 8.5; 9.1; 10.1 and 11.2 groupings are the same in all three research, while 7.3 was found only in <20 and 20 scoring alcoholics. I can suggest that this way of grouping belongs only to surely demented people, according to Bertolani, De Renzi & Faglioni, 1993. So I can also suggest that grouping 9.5 found only in 20 and 21-25 scoring people belongs to dubious demented people.

On the other hand, groupings 4.4; 7.1; 10.2 and 11.3 found both to <20 and 21-25 scoring people probably lack the same groupings in the present sample because its small number of subjects.

Discussion.

To divide the wrong answers, the categories in use came out from a research on college students, using PM47, Revision 1962, Form I (Cocchi, 1993). In the previous two papers on RCM in alcoholics scoring less than 20 (Cocchi 1963), and between 21 and 25 (Cocchi, 1996), all wrong answers took place into five categories, but only few of them into "globality (or, totality) answer" and none into "first not present" answers. The third paper on the same topic (Cocchi and Meme', 1998) needed only four categories for categorization, without any "globality (or, totality) answer".

Although the present research deals with scoring exactly between the first and the second paper, "globality (or, totality) answer " and "similarity" fields had not any room,

In previous research in alcoholics scoring < 20 (Cocchi, 1993) or 21-25 (Cocchi, 1996), I noted "globality answer" as a stop signal in processing the choice.

The patient makes of how the matrix fills, but s/he seems unable to leave this total image behind. So s/he does not switch to the second step, by choosing what fills in, instead of the wrong answer showing the filled matrix.

Perhaps this results because there were only 20 probands and related figures do not reach any statistical signification. The same explanation could work for the lacking of any "similarity" answer.

"Confabulation" as a spreding field needs to be better specified. With this term I think of an answer where out context and plain elements join on.

But this type of wrong answer is not always at ease to find, because only few tables allow of such a choice. A "confabulation" answer works as partial identity, a metaphor, but the outside element identifies it with the same term used in speech pathology.

In this research confabulation takes the third place, while it was quarter in alcoholics who scored less than 20 at RCM (Cocchi, 1993). In college students also I found confabulation just after contiguity and opposition (Cocchi, 1993).

As for the remaining two categories (i.e.: contiguity and opposition), I am referring again to what I wrote in my previous papers (Cocchi, 1993, Cocchi 1996; Cocchi and Meme' 1998). I considered them, together with similarity, the result of brain awareness of its neuronal network functioning. I did a detailed discussion of it in my paper of the analysis of possible neurofisiological bases of figures of speech (Cocchi 1997).

Comparing wrong answers groupings of people scoring <20, 20 and 21-25 we can see an increased fugure of the same wrong answers' groupings according RCM tables difficulties (from A to B). On the other hand the small number of subjects of the 20 scoring sample dealt to not reach significantly groupings otherwise found both in <20 and in 21-25 scoring samples. Other hsuggestions could be forwarded but I think now this is enough.

The analysis of 316 wrong answers at RCM 20 alcoholics scoring 20 did first days of their inpatient stay let out 89 wrong answers on 21 non random groups. All these non random wrong answers were split into three categories such as contiguity, opposition, confabulation, in reduced amounts.

Some ways of neural network functioning can account for that.

The results confirm again that wrong answers given at RCM can notice the neuronal levels where alcoholics stop reasoning in a problem solving task. Twenty scoring Ss stand between <20 and 21-25 scoring samples with an evident trend closer to the 21-25 scoring sample.

Bertolani L., De Renzi E., Faglioni P.: Test di memoria non verbale di impiego diagnostico in clinica: Taratura su soggetti normali. Arch. Psicol. Neurol. Psichiat. 1993, 54: 477-486.

Cocchi R.: Analisi delle risposte errate, date alle PM47 di Raven, Rev. 1962, Forma I, da un campione di studenti universitari. Riv. Ital. Disturbo Intellet. 1993, 6: 83-90.

Cocchi R.: Alcolisti con punteggio < 20 alle Matrici Colorate di Raven: Analisi degli errori. Riv. Ital. Disturbo Intellet. 1993, 6: 269-275.

Cocchi R.: Alcoholics scoring 21-25 at Raven's Coloured Matrices: The analysis of wrong answers. It. J. Intellect Impair. 1996, 9: 181-187.

Cocchi R.: Figures of speech: Possible neurophysiological bases? It. J. Intellect Impair. 1997, 10: 35-42.

Cocchi R., Meme' M.: Alcoholics scoring 26-35 at Raven's Coloured Matrices: The analysis of wrong answers It. J. Intellect. Impair. 1998, 11: 173-178.

Pola A., Cocchi R., Zerbi F.: Progressive Matrices PM 47 in demented inpatients: qualitative analysis of mistakes and problem solving strategies. Ital. J. Intellect. Impair. 1988, 1: 111-118.

Author's address: dr Renato COCCHI, via Rabbeno, 3

42100 Reggio Emilia.